She didn’t “botch” a study.
She made shit up to suit the Narrative.
She better have her Waffle House job application filled out.
I would suggest all those years in high school and college where she got a ‘free-ride’...reached a level where she finally had to produce a real study, and she couldn’t do it. As for the college now admitting her limits or their mistake? No...to fire her means a lawsuit and tens of millions to be paid for firing. She has a 30-year job in her hands and will never leave.
She actually was inconsistent. In one paper she showed one value for the statistic and standard error, in another different values. The second paper, probably the bogus one, indicates a significant decrease in white Democratic voter turnout when the Democrats offer a black candidate, the first one showed no significant effect, both “statics”, white voter turnout and standard error were supposedly analysis of the exact same election.
The original controversy arose in 2002 when other researchers asked to see the underlying data for the second paper. She used a somewhat exotic technique, which was not necessarily invalid or unreliable.
Since ballots are not (yet) marked with voter’s race, she used turnout in white precincts as a proxy for white voter turn out. Within limits one can make rough estimates of turnout based on these data. For instance, it is widely believed that black turnout for Barack Obama was significantly greater among black voters than in other elections where there was not a black (or at least Halfrican) candidate for president, based on turn out in black precincts nationwide.